[go: up one dir, main page]

WO2017002812A1 - Dispositif de gestion d'infrastructure de virtualisation, dispositif de gestion de fonction de réseau virtuel, procédé de gestion de machine virtuelle et programme - Google Patents

Dispositif de gestion d'infrastructure de virtualisation, dispositif de gestion de fonction de réseau virtuel, procédé de gestion de machine virtuelle et programme Download PDF

Info

Publication number
WO2017002812A1
WO2017002812A1 PCT/JP2016/069149 JP2016069149W WO2017002812A1 WO 2017002812 A1 WO2017002812 A1 WO 2017002812A1 JP 2016069149 W JP2016069149 W JP 2016069149W WO 2017002812 A1 WO2017002812 A1 WO 2017002812A1
Authority
WO
WIPO (PCT)
Prior art keywords
virtual
network function
virtualization infrastructure
function virtualization
virtual machine
Prior art date
Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
Ceased
Application number
PCT/JP2016/069149
Other languages
English (en)
Japanese (ja)
Inventor
雅知 椋毛
淳一 極樂寺
博一 篠澤
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
NEC Corp
Original Assignee
NEC Corp
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by NEC Corp filed Critical NEC Corp
Publication of WO2017002812A1 publication Critical patent/WO2017002812A1/fr
Anticipated expiration legal-status Critical
Ceased legal-status Critical Current

Links

Images

Classifications

    • G—PHYSICS
    • G06—COMPUTING OR CALCULATING; COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F9/00—Arrangements for program control, e.g. control units
    • G06F9/06—Arrangements for program control, e.g. control units using stored programs, i.e. using an internal store of processing equipment to receive or retain programs
    • G06F9/46—Multiprogramming arrangements
    • G—PHYSICS
    • G06—COMPUTING OR CALCULATING; COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F9/00—Arrangements for program control, e.g. control units
    • G06F9/06—Arrangements for program control, e.g. control units using stored programs, i.e. using an internal store of processing equipment to receive or retain programs
    • G06F9/46—Multiprogramming arrangements
    • G06F9/50—Allocation of resources, e.g. of the central processing unit [CPU]

Definitions

  • the present invention is based on the priority claim of Japanese patent application: Japanese Patent Application No. 2015-130722 (filed on June 30, 2015), the entire description of which is incorporated herein by reference. Shall.
  • the present invention relates to a virtual infrastructure management apparatus, a virtual network function management apparatus, a virtual machine management method, and a program, and more particularly to management of a virtual machine used to provide a virtual network function.
  • Virtualization technology makes it possible to use a single physical machine resource in multiple virtual machines (hereinafter referred to as “VM”).
  • VM virtual machines
  • the resources of the physical machine are limited, if one VM occupies the resource, resources necessary for stable operation of the other VM cannot be secured, processing delay occurs, and congestion / A failure may occur.
  • Patent Document 1 in a virtual machine system in which a plurality of virtual machines are operated using a plurality of physical machines, when a shortage of resources occurs in any physical machine, the shortage of resources can be appropriately resolved.
  • a virtual machine management device capable of being disclosed is disclosed. According to this document, when a resource shortage occurs, this virtual machine management apparatus identifies a physical machine having a remaining resource amount equal to or more than a predetermined amount as a specific physical machine.
  • the virtual machine management device 5 uses the resource usage of the virtual machine that operates on the physical machine in which the resource shortage occurs to cause resource contention when operating on the specific physical machine from the virtual machine that operates on the physical machine. Identify a virtual machine that is unlikely to generate as a specific virtual machine. Then, it is described that the virtual machine management device 5 moves the physical machine that operates the specific virtual machine to the specific physical machine.
  • Patent Document 2 discloses a technology for scaling out a virtual server. According to this document, the monitoring server 4 detects the virtual server that caused the high load, acquires a system that depends on the detected virtual server process, and acquires only the acquired system. It is described that the configuration change server 6 moves the virtual server to the scale-out destination.
  • Non-Patent Document 1 discloses network function virtualization (hereinafter referred to as “NFV”) that realizes functions of various communication devices on a network on a virtualized operating system (OS) of a general-purpose server using the above-described virtualization technology. ”)" Is disclosed.
  • the NFV is realized based on, for example, a MANO (Management & Orchestration) architecture.
  • FIG. 7 is a figure cited from FIG. 5.1 (The NFV-MANO architectural framework with reference points) on page 23 of Non-Patent Document 1.
  • VNFM Virtualized Network Function Manager
  • VNFM Virtualized Network Function Manager
  • FIG. 8 is a diagram for explaining a case where the inappropriate healing target is selected.
  • the upper part of FIG. 8 shows a state where the CPU core 1 of the physical server of the network function virtualization infrastructure (hereinafter “NFVI”) is assigned to VM1, and the CPU core 2 is assigned to VM2 and VM3.
  • NFVI network function virtualization infrastructure
  • Patent Document 1 when a plurality of VMs share the same physical resource, a VM (virtual machine memory usage / CPU usage rate, frequency, which is unlikely to cause resource contention in the VM to be healed) Is low VM) (see paragraph 0029).
  • a VM virtual machine memory usage / CPU usage rate, frequency, which is unlikely to cause resource contention in the VM to be healed
  • a similar situation can occur with (auto) scaling. That is, if the (auto) scaling condition is satisfied, a VM that has been operating stably may be selected as a target for scaling or NFVI movement.
  • the present invention provides a virtual infrastructure management apparatus, a virtual network function management apparatus, a virtual machine management method, and a program capable of contributing to the suppression of a service that has been stably operated being interrupted by the healing or scaling described above.
  • the purpose is to provide.
  • the virtual machine includes a VM monitoring unit that acquires a resource usage status of a virtual machine that operates while sharing a physical resource on the first network function virtualization infrastructure, and manages it as statistical information.
  • a virtual infrastructure management apparatus is provided.
  • the virtualization infrastructure management apparatus further includes the event that an event of moving a virtual machine operating on the first network function virtualization infrastructure to the second network function virtualization infrastructure occurs. Based on the statistical information, one or more virtual machines are selected from the virtual machines operating while sharing physical resources, and the second network function virtualization is selected from the first network function virtualization infrastructure.
  • a VM migration control unit that migrates to the infrastructure.
  • a VM monitoring unit that acquires the resource usage status of a virtual machine that operates while sharing physical resources on the first network function virtualization infrastructure, and manages the information as statistical information; When an event for moving a virtual machine operating on the first network function virtualization infrastructure to the second network function virtualization infrastructure occurs, physical resources are shared based on the statistical information.
  • a VM migration control unit that selects one or more virtual machines from the operating virtual machines and moves the virtual machines from the first network function virtualization infrastructure to the second network function virtualization infrastructure; Virtualization infrastructure with In response to a service request from the user and the structure management unit, the virtual network function necessary for providing the service is determined, and the virtual infrastructure management unit is instructed to create a virtual machine corresponding to the virtual network function.
  • a virtual network function management device including an orchestration unit.
  • an event occurs to move a virtual machine operating on one network function virtualization infrastructure to the second network function virtualization infrastructure, it operates by sharing physical resources based on the statistical information Selecting one or more virtual machines from among the virtual machines being moved from the first network function virtualization infrastructure to the second network function virtualization infrastructure.
  • Management method provided It is. This method is linked to a specific machine called the virtual infrastructure management apparatus described above.
  • a computer program for realizing the functions of the virtual infrastructure management apparatus or the virtual network function management apparatus described above is provided.
  • This program can be recorded on a computer-readable (non-transient) storage medium. That is, the present invention can be embodied as a computer program product.
  • the present invention it is possible to contribute to the prevention of interruption of a service that has been operating stably due to the healing and scaling described above. That is, the present invention converts the virtual infrastructure management apparatus shown in the background art into a virtual infrastructure management apparatus with improved service continuity.
  • the present invention can be realized by a virtual infrastructure management apparatus 10 including a VM monitoring unit 11 and a VM migration control unit 12, as shown in FIG.
  • the VM monitoring unit acquires the resource usage status of the virtual machine operating on the first NFVI 1 while sharing the physical resource, and manages it as statistical information.
  • the VM movement control unit 12 When an event for moving a VM operating on NFVI1, such as healing or scaling, to NFVI2 occurs, the VM movement control unit 12 is configured to share a physical resource based on the statistical information. One or more virtual machines in which an abnormality has occurred are selected, and moved from the NFVI1 to the NFVI2.
  • the VM migration control unit 12 selects the VM 2 from the VM 2 and VM 3 that operate while sharing the physical resource (CPU core 2), moves the VM 2 from the NFVI 1 to the NFVI 2, Can be scaled out to VM2-2. By doing so, it is possible to continue the operation of the VM 3 that has been operating without any problems while performing the scale-up and to continue the service.
  • a method for selecting a VM having an abnormal state from VMs operating while sharing physical resources For example, a method of selecting a VM whose value (CPU usage rate) related to the physical resource (CPU core) to be shared is higher than a predetermined threshold among the values monitored as the statistical information is adopted. Can do. Further, in combination with the comparison of the threshold values, VMs of the same user or the same virtual network may be selected as a movement target.
  • FIG. 3 is a diagram illustrating a configuration of the virtual infrastructure management apparatus (VIM) according to the first embodiment of this invention.
  • the configuration includes NFVI1, three VM1 to VM3 operating on NFVI1, and a VIM 100 that controls resource allocation and the like to the VM in response to an instantiation (service request) of the VNF from the host device. It is shown.
  • the VIM 100 includes a controller 101 that acquires the usage status of virtual machine resources that operate while sharing physical resources on the first NFVI 1 and manages them as statistical information. Assume that the controller 101 monitors the CPU usage rate / memory / storage / NW disk I / O rate of each VM as statistical information. Note that some of these may not be actively monitored by the controller 101, but may be configured to notify the controller 101 from VM or VNF periodically or when an abnormality occurs.
  • the NFV-MANO configuration shown in FIG. 7 can be adopted as the other basic configuration. That is, the VIM 100 operates in cooperation with the orchestrator shown in FIG. 7 to provide VNFs according to the user service request. Further, the VIM 100 has a function of selecting a VM to be healed or scaled based on statistical information acquired from the controller 101.
  • VM1 and VM3 are in a state of sharing a resource called CPU core 2.
  • CPU core 2 is assigned to VM2 and VM3.
  • VM2 and VM3 are in a state of sharing a resource called CPU core 2.
  • the CPU core is taken up as a resource, there are cases where a memory, a disk, etc. are shared.
  • VM1 to VM3 VM1 and VM2 belong to virtual network 1 (vLAN1), and VM3 belongs to another virtual network 2 (vLAN2) (see the lower dotted line in FIG. 3).
  • the virtual infrastructure management apparatus or each unit (processing means) of the VIM shown in FIGS. 1 and 3 is a computer that causes a computer constituting these apparatuses to execute the above-described processes using the hardware. It can also be realized by a program.
  • FIG. 4 is a flowchart showing the operation of the virtual infrastructure management apparatus when healing (or scaling) occurs as an example of an event of moving a VM operating on a certain NFVI to another NFVI.
  • the VIM 100 refers to its own VNF flavor (see 6.3.1.5 Deployment flavour in Non-Patent Document 1), and multiple VMs are the same It is determined whether the NFVI resource is shared (step S001).
  • the nonpatent literature 1 can be used for the execution condition (starting condition) of healing (or scaling).
  • the healing may include not only auto-healing but also manual healing or scaling by operator operation as described in Non-Patent Document 1.
  • the VIM 100 acquires the CPU usage rate / memory / storage / NW disk I / O rate of each VM as statistical information from the controller 101. Then, by comparing with the threshold value determined for each, a VM having an abnormal state is identified and selected as a VM for healing / scaling (step S002).
  • the VIM 100 After the VM to be healed (or scaled) is specified, the VIM 100 performs healing (or scaling) on the corresponding VM.
  • step S002 a method for selecting a VM to be healed (or scaled) in step S002 will be described.
  • the usage of the CPU core 2 to which the VM 2 is allocated has risen due to congestion, the occurrence of a failure, etc. in the VM 2, and the threshold has been exceeded.
  • the description will be made assuming that the healing condition is satisfied.
  • the VIM 100 identifies the healing (or scaling) VM from the statistical information acquired from the controller 101.
  • the VMM 100 refers to the CPU usage rates of the VM2 and VM3 sharing the same CPU core 2, and detects that the VM usage rate of the VM2 has risen and that the VM2 is congested or abnormal. On the other hand, since there is no significant change in the CPU usage rate for VM3, it is in a normal state.
  • the VIM 100 performs healing for only the selected VM2.
  • the VIM 100 starts up VM2 and performs healing in another NFVI2 based on the VM image in the VNF catalog according to the instruction from the orchestrator of FIG. Since VM3 is not subject to healing, the service is continued at the VNF corresponding to VM3.
  • a plurality of VMs use the same NFVI resource, it is possible to select only a VM having a state abnormality and make it a target for healing (or scaling). As a result, forced healing (or scaling) of other VMs that are operating stably can be prevented, and service availability can be ensured.
  • the VM CPU usage rate / memory / storage / network disk I / O rate and the like are compared with respective threshold values, and the presence / absence of a VM abnormality is determined.
  • a state in which an arbitrary value of statistical information (for example, CPU usage rate) exceeds a threshold th1 is determined to be “abnormal” when the state continues for a predetermined period. (Case 1 in FIG. 6).
  • the presence or absence of VM abnormality may be comprehensively determined with reference to a plurality of values of statistical information. For example, as in the case of VM2 in FIG. 6, even if an arbitrary value of statistical information (for example, CPU usage rate) does not exceed the threshold th1, other values of statistical information indicate high values. If there is, it may be determined as “abnormal” (case 2 in FIG. 6).
  • a service is provided on the same virtual network or a service is provided to the same user.
  • the VM 1 in FIG. 6 is the object of healing (or scaling)
  • the VM 3 that provides a service in the same virtual network may be added to the object of healing (or scaling) (case 3 in FIG. 6). This is because in the case of VMs that provide services on the same virtual network, it is considered that the impact on user availability is not significant. Based on this concept, it may be possible to perform healing (or scaling) in units of VMs that provide services in the same virtual network, rather than whether or not an abnormality has occurred.
  • the present invention can be applied to a case where a VM is stopped and restarted when a predetermined condition is satisfied or the user explicitly operates. For example, when a VM using a certain physical resource is in an automatic stop state for some reason and then restarted, it is moved from the VM sharing the physical resource to another NFVI A VM may be selected.
  • the VM determined to be “abnormal” from the VMs sharing the physical resource has been described as being moved.
  • a VM that is not determined to be “abnormal” may be moved.
  • This restriction is not limited to a hardware restriction.
  • a case where a restriction is added to NFVI that can be used by the corresponding VM user by contract or the like is also included.
  • the VM migration control unit selects a virtual machine whose resource usage exceeds a predetermined threshold among virtual machines that operate while sharing physical resources, as a migration target.
  • Infrastructure management device In the virtual infrastructure management apparatus of the first or second form, The VM migration control unit is a virtualization infrastructure management apparatus that further selects a virtual machine that operates on the same virtual network among virtual machines that operate while sharing physical resources.
  • a virtual infrastructure management device that operates in cooperation with an orchestrator of a virtual network function and provides a virtual network function according to a user service request.
  • the event is the establishment of a scale-out condition
  • the VM migration control unit is a virtualization infrastructure management apparatus which is a scale-out process for starting a plurality of virtual machines operating in the first network function virtualization infrastructure in the second network function virtualization infrastructure.
  • the event is the establishment of a virtual machine healing condition
  • the VM movement control unit A virtualization infrastructure management apparatus, which is a healing process for reconfiguring a virtual machine operating in a first network function virtualization infrastructure with the second network function virtualization infrastructure.
  • the virtual infrastructure management apparatus wherein the statistical information is at least one statistical information of a virtual CPU usage rate, a virtual memory usage rate, a virtual storage usage rate, and a virtual disk IO rate of a virtual machine.

Landscapes

  • Engineering & Computer Science (AREA)
  • Software Systems (AREA)
  • Theoretical Computer Science (AREA)
  • Physics & Mathematics (AREA)
  • General Engineering & Computer Science (AREA)
  • General Physics & Mathematics (AREA)
  • Data Exchanges In Wide-Area Networks (AREA)

Abstract

La présente invention a pour but de réduire les interruptions de services qui fonctionnaient de manière stable, dues à un rétablissement ou une mise à l'échelle lors de la fourniture d'une fonction de réseau virtualisée (VNF). La présente invention concerne un dispositif de gestion d'infrastructure de virtualisation, comprenant : une unité de surveillance de machine virtuelle (VM) qui acquiert une situation d'utilisation de ressources de machines virtuelles qui fonctionnent avec des ressources physiques partagées sur une première infrastructure de virtualisation de fonction de réseau (NFVI), et gère lesdites situations d'utilisation en tant qu'informations statistiques ; et une unité de commande de déplacement de VM qui, si un événement s'est produit, dans lequel les machines virtuelles qui fonctionnent sur la première infrastructure de virtualisation de fonction de réseau sont déplacées vers une seconde infrastructure de virtualisation de fonction de réseau, sélectionne, sur la base des informations statistiques, une ou plusieurs des machines virtuelles parmi les machines virtuelles qui fonctionnent avec les ressources physiques partagées, et déplace les machines virtuelles sélectionnées de la première infrastructure de virtualisation de fonction de réseau à la seconde infrastructure de virtualisation de fonction de réseau.
PCT/JP2016/069149 2015-06-30 2016-06-28 Dispositif de gestion d'infrastructure de virtualisation, dispositif de gestion de fonction de réseau virtuel, procédé de gestion de machine virtuelle et programme Ceased WO2017002812A1 (fr)

Applications Claiming Priority (2)

Application Number Priority Date Filing Date Title
JP2015130722 2015-06-30
JP2015-130722 2015-06-30

Publications (1)

Publication Number Publication Date
WO2017002812A1 true WO2017002812A1 (fr) 2017-01-05

Family

ID=57609513

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/JP2016/069149 Ceased WO2017002812A1 (fr) 2015-06-30 2016-06-28 Dispositif de gestion d'infrastructure de virtualisation, dispositif de gestion de fonction de réseau virtuel, procédé de gestion de machine virtuelle et programme

Country Status (1)

Country Link
WO (1) WO2017002812A1 (fr)

Cited By (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN108628660A (zh) * 2017-03-24 2018-10-09 华为技术有限公司 一种虚拟机扩缩容方法及虚拟管理设备
WO2020261412A1 (fr) * 2019-06-26 2020-12-30 日本電信電話株式会社 Dispositif de commande basée sur une virtualisation, procédé de commande basée sur une virtualisation et programme de commande basée sur une virtualisation

Citations (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
WO2012066640A1 (fr) * 2010-11-16 2012-05-24 株式会社日立製作所 Système informatique, procédé de migration et serveur de gestion
WO2014030203A1 (fr) * 2012-08-20 2014-02-27 富士通株式会社 Dispositif de gestion d'opérations, procédé de gestion d'opérations et programme de gestion d'opérations
US20140201374A1 (en) * 2013-01-11 2014-07-17 Futurewei Technologies, Inc. Network Function Virtualization for a Network Device
JP2014167713A (ja) * 2013-02-28 2014-09-11 Fujitsu Ltd 情報処理装置、情報処理システム、情報処理装置管理プログラム及び情報処理装置管理方法

Patent Citations (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
WO2012066640A1 (fr) * 2010-11-16 2012-05-24 株式会社日立製作所 Système informatique, procédé de migration et serveur de gestion
WO2014030203A1 (fr) * 2012-08-20 2014-02-27 富士通株式会社 Dispositif de gestion d'opérations, procédé de gestion d'opérations et programme de gestion d'opérations
US20140201374A1 (en) * 2013-01-11 2014-07-17 Futurewei Technologies, Inc. Network Function Virtualization for a Network Device
JP2014167713A (ja) * 2013-02-28 2014-09-11 Fujitsu Ltd 情報処理装置、情報処理システム、情報処理装置管理プログラム及び情報処理装置管理方法

Cited By (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN108628660A (zh) * 2017-03-24 2018-10-09 华为技术有限公司 一种虚拟机扩缩容方法及虚拟管理设备
WO2020261412A1 (fr) * 2019-06-26 2020-12-30 日本電信電話株式会社 Dispositif de commande basée sur une virtualisation, procédé de commande basée sur une virtualisation et programme de commande basée sur une virtualisation
JPWO2020261412A1 (fr) * 2019-06-26 2020-12-30
JP7176633B2 (ja) 2019-06-26 2022-11-22 日本電信電話株式会社 仮想化基盤制御装置、仮想化基盤制御方法および仮想化基盤制御プログラム

Similar Documents

Publication Publication Date Title
EP3468151B1 (fr) Procédé et appareil de traitement de ressources d'accélération
US10609159B2 (en) Providing higher workload resiliency in clustered systems based on health heuristics
US20150309828A1 (en) Hypervisor manager for virtual machine management
CN105051698B (zh) 用于基础设施即服务云中故障管理的方法和布置
KR101939075B1 (ko) 시스템, 가상화 제어 장치, 가상화 제어 장치의 제어 방법 및 프로그램
US9875145B2 (en) Load based dynamic resource sets
US10846119B2 (en) Virtualized network function management apparatus, virtual machine management apparatus, method for allocating resources to virtual network function, and program
CN107026802A (zh) 一种资源管理方法及设备
WO2015101091A1 (fr) Procédé et dispositif d'ordonnancement distribué de ressources
US10228963B2 (en) Efficient hypervisor reporting
US10108463B2 (en) System, method of controlling to execute a job, and apparatus
CN108028806B (zh) 网络功能虚拟化nfv网络中分配虚拟资源的方法和装置
KR20130019698A (ko) 사용자 스케줄러와 마이그레이션(Ｍｉｇｒａｔｉｏｎ)을 통한 자원 최적화 방법 및 시스템
JP2018180591A (ja) 情報処理装置、情報処理システム、情報処理方法及び情報処理プログラム
EP2645635B1 (fr) Moniteur de grappe, procédé permettant de surveiller une grappe et support d'enregistrement lisible par ordinateur
JP6993495B2 (ja) クラウド・ネットワーキングにおけるスケーラブルな統計及び分析メカニズム
KR102010152B1 (ko) 가상머신의 마이그레이션을 수행하기 위한 방법 및 장치
WO2017002812A1 (fr) Dispositif de gestion d'infrastructure de virtualisation, dispositif de gestion de fonction de réseau virtuel, procédé de gestion de machine virtuelle et programme
EP3151513A1 (fr) Procédé et dispositif élastiques de service en informatique en nuage
US10616064B2 (en) Soft reservation techniques and systems for virtualized environments
WO2017018435A1 (fr) Dispositif de surveillance de ressources, système de gestion de fonction de réseau virtuel, procédé de surveillance de ressources, et programme associé
TWI522921B (zh) 虛擬機器建立系統以及方法
CN107562510B (zh) 一种应用实例的管理方法及管理设备
KR101737468B1 (ko) 가상화 환경에서의 자원 관리 장치 및 방법
JP2015121963A (ja) 情報処理システム、監視方法、及び、プログラム

Legal Events

Date Code Title Description
121 Ep: the epo has been informed by wipo that ep was designated in this application

Ref document number: 16817920

Country of ref document: EP

Kind code of ref document: A1

NENP Non-entry into the national phase

Ref country code: DE

122 Ep: pct application non-entry in european phase

Ref document number: 16817920

Country of ref document: EP

Kind code of ref document: A1

NENP Non-entry into the national phase

Ref country code: JP